zhongcao-cover-maker
Pass
Audited by Gen Agent Trust Hub on Nov 14, 2026
Risk Level: SAFEREMOTE_CODE_EXECUTIONCREDENTIALS_UNSAFEDATA_EXFILTRATIONINDIRECT_PROMPT_INJECTIONEXTERNAL_DOWNLOADSCOMMAND_EXECUTION
Full Analysis
- [REMOTE_CODE_EXECUTION]: The installation process involves cloning a GitHub repository and running Python scripts such as
authorize.pyandmcp_client.py. - [CREDENTIALS_UNSAFE]: The skill stores a bearer token with financial scope (
wallet:spend) in the local file~/.beatra/credentials.json. - [DATA_EXFILTRATION]: The skill transmits user prompts, uploaded media, and system telemetry including hostname, OS version, CPU architecture, and locale to external endpoints at
api.beatra.aiandapi.beatra.cn. - [INDIRECT_PROMPT_INJECTION]: The skill processes untrusted user-supplied photos and topic ideas for processing by a third-party service. Ingestion points: User-supplied images and text descriptions. Boundary markers: None specified in the documentation to prevent instruction override in processed media. Capability inventory: Network requests to external APIs and local file system modifications. Sanitization: No input validation or sanitization of external content is mentioned.
- [EXTERNAL_DOWNLOADS]: The skill fetches its core execution components from a third-party GitHub repository at installation time.
- [COMMAND_EXECUTION]: The installation instructions require the user to run shell commands to configure the environment, review the tree, and copy files to the agent's directory.
Audit Metadata